Atrial fibrillation (AF) is an arrhythmia with complex and not well understood mechanisms which make treatment of AF a significant challenge. Multiple factors likely contribute to whether a person is susceptible to AF, including the existence of focal triggers, the complex anatomy of the left atrium, and the modification of atria substrate (e.g. fibrosis). Cardiac imaging, through MRI or CT, has been used to evaluate the three- dimension structure of the atria as guidance for radiofrequency ablation of AF. More recently, it has been shown that MRI can be used to detect fibrosis and ablation lesions in the atria. Despite these advances in atrial imaging, it is not clear how these data can be used to evaluate or treat of AF. We propose a novel approach to evaluate atrial geometry and substrate, as assessed by cardiac MRI, through computer simulations where realistic electrophysiologic properties are assigned to the three-dimensional geometry and by which "virtual" AF induction can be performed. The first aim of the project is to obtain the optimal parameters for obtaining high quality delayed-enhancement MRIs that will allow delineation of the atrial wall and detection of structural remodeling with reasonable scan times. In the second aim, a simulation tool will be developed to segment the atria and to apply a threshold-based algorithm to characterize each voxel of the atrial wall as being normal, structurally remodeled but viable, or structurally remodeled and non-viable. The tool will convert the three- dimensional voxel data into a monolayer polygon model on which computer simulations will be performed. A computer model that realistically simulates atrial electrophysiology will be used. The third aim will test the hypothesis that computer simulations incorporating MRIs taken prior to ablation and three months after ablation can be used to predict likelihood of success for ablation. The reason for the recurrence will be determined by comparing the computer simulations using the post-ablation MRI against the computer simulation using the pre-ablation MRI with the intended ablation plan. Incomplete ablation lesions or an inadequate ablation strategy may be implicated as reasons for the recurrence. This project represents a novel approach to study computer simulations of AF with MRI data from a large set of patients and is the first to test computer modeling as a potential clinical tool to evaluate AF patients. This study could have significant implications on the development of individualized strategies for radiofrequency ablation to treat AF. PUBLIC HEALTH RELEVANCE: Atrial fibrillation is rapidly emerging health issue in the United States, particularly among the elderly. Without the clear understanding of the mechanisms of atrial fibrillation, treatment options, such as radiofrequency ablation, remain suboptimal in terms of efficacy and risks.